Review History
Analysis of the genetic diversity of the coastal and island endangered plant species Elaeagnus macrophylla via conserved DNA-derived polymorphism marker

Reviewer 3 · Oct 25, 2019

Additional comments

Lines 124-125
Replace:
… and gene flow (Nm=1-Gst/Gst) among populations or localities when significant differences among specimens were found.
By:
… and gene flow (Nm=1-Gst/Gst) among populations. The estimates were also calculated within sampling localities when significant differences among specimens were found.


Lines 134-137
Please, revise this statement considering its grammatical improvement
Compress the result of the operation (".zip" format) and upload it to the Structure Harvester website (http://ztaylor0.bi-ology.ucla.edu/structure) to get the analysis result of the K value (Evanno, Regnaut & Goudet, 2005.)


Lines 227-230
Replace:
Frankham compared and analyzed the allelic diversity of 202 groups of land and island
populations of various species, including mammals, birds, fish, reptiles, insects and plants; in
165 cases (81.7%), the genetic diversity of island populations was lower than that of terrestrial
populations, with an average decrease of 29% (Frankham, 1997)
By:
Frankham (1997) compared and analyzed the allelic diversity of 202 groups of land and island
populations of various species, including mammals, birds, fish, reptiles, insects and plants; in
165 cases (81.7%), the genetic diversity of island populations was lower than that of terrestrial
populations, with an average decrease of 29%.


Line 288
… clustering analysis shows that there was no hybridization among individuals
Isn’t it supposed to be “there was no hybridization among individuals from different localities”?


About Delta-K values
Despite authors have answered in rebuttal letter that the Delta-K values were provided in supplemental files, I believe it could be presented within Figure 5 (e.g., Fig. 5B). This result is important to prove how good are the model chosen by authors after Bayesian analysis.
Besides, I also have tow additional comments/questions about Structure results:
a) It’s clear that Structure pointed that E. macrophylla genetic diversity can be arranged into three main genetic groups. There was also a substructure found considering 5 genetic groups. Authors described very well the importance of the 5 subgroups. However, in discussion, the 3 main genetic groups were neglected. What are the main conclusions about these 3 main genetic groups?
b) The occurrence of subgroups in NJD population was well described by authors. They also repeated some analysis considering only this samples (line 216; Gst and Nm of the two sub-populations of NJD). However, its clear for me that they also have another subpopulation structure in DRS samples. It can be observed in Figure 5 and, also, in the dendrogram. Part of this specific result for DRS is mentioned briefly in results, while in NJD we have an specific section to show this findings (line 216; Gst and Nm of the two sub-populations of NJD). A discussion about this result is also absent in the manuscript.

Reviewer 3 · Sep 25, 2019

Additional comments

a) Authors must check the PeerJ instructions for citing references. There are some references cited (mainly in discussion) that doesn’t fit the instructions. E.g.:
•Line 236: “Weller et al., 1996”; According to the journal rules, this paper has to be cited as “Weller, Sakay & Straub, 1996”. Please, check all the “three-authors” references.
•Lines 314-316: “Leng's research also mentioned that when the seeds of Ilex integra were soaked in seawater, the germination rate was approximately zero (Leng et al., 2005).” This isn’t the most appropriate way to cite a reference. I suggest something like “Leng et al. (1995) also mentioned that when the seeds of Ilex integra were soaked in seawater, the germination rate was approximately zero.” … or … “Seeds of Ilex integra also presented germination rate close to zero when soaked in seawater (Leng et al., 2005).” There are several references cited in this way within the Discussion.

b) Lines 77-82: This statement is too long. Breaking the sentence may provide a clearer information. Also, correct the “of and” at the end of the line 77.

c) Lines 132-134: Structure Harvester implements the method described by Evanno et al. (2005; doi: 10.1111/j.1365-294X.2005.02553.x). Authors have to insert this reference in the text since they have used the method.

d) Lines 162-163: This is a “discussion” statement and should not be descried within “results” section.

e) Lines 180-195: It's still not clear which criteria the authors used to split the samples in the different groups. I can see that the “considered groups” are based in the main branches observed in the dendrogram. However, this is a little arbitrary. Are these "groups" consistent? Based on what the authors know that they really have “true groups” in this analysis? Why not presenting the bootstrap to improve the statement? If they used the "cut-line" (e.g., based on the average similarity), it is not showed in Figure 4, neither in the “Material and Methods” section. It's important to make it clear, so the reader will know the basis of the method used to split the samples in different groups. Another way to avoid this problem is bringing the focus of the statement for the fact that no cross-clustering were observed among samples of different locations. Regardless of this information is present in lines 189-190, its sounds like a “secondary” information in the text.

f) Lines 199-214: Considering that authors have used the Evanno et al. (2005) criteria (delta K) to detect the most probable number of clusters, the results of delta K must be showed in the manuscript. Authors showed values for K=3 and K=5. But I cannot evaluate if these solutions are the best ones for the samples since I don’t have the magnitude of delta-K values. Please, provide delta-K values or the graph showing the likelyhood for each tested K value. E. g., authors said that "when K = 5, the value of delta K is also large”. But, how large is it?

g) Line 248: “delicious” is a very relative terminology. I’m not sure if Zang (2016) have really used this word but, in general, it is not well suited in scientific language.

h) Line 330: It’s interesting that Gst values indicates that NJD can be arranged in two subpopulations. However, this information was not presented in the Results section. It is also important to mention this approach in Methods, saying that the Gst index were calculated within localities when significant differences among specimens were found.

i) Include the DOI in the references listed.

Reviewer 1 · Jul 26, 2019

Additional comments

Article must be rechecked thoroughly and added more scientific contents to make it a relevant study for the readers.

Reviewer 2 · Jul 29, 2019

Reviewer 3 · Aug 2, 2019

Additional comments

1. Introduction: readers need more information concerning biological issues of E. macrophylla. Ecological and reproductive information of the species are needed to understand the results. Whats the mating system? Whats the main pollinators? How does the dispersion happen? How well this plant is fitted to the coastal/island ambient? Its important to know if a plant uses self or cross-fertilization to understand the diversity data based on molecular makers. Selfing may naturally decreases the magnitude of indexes like heterozygosity in plants and, this decrease may not necessarily represent a poor diversity for some species. So its important to give the elementary biological information of the species for readers understand your results.

2. There are some questions about CDDP markers that must be addressed. As other papers suggested, authors justifies the use of this maker due to its reproducibility when compared with other dominant markers (e.g., RAPD and ISSR). However, even the CDDP developers (Collard and Mackill, 2009) warned that reproducibility problems also exists for CDDP, “suggesting that primer length and annealing temperature per se do not ensure reproducibility”. They recommend gel scoring based on at least two independent replicates of PCRs. It is not clear if authors were careful enough considering the reproducibility problems. Reading the text, there are no elements discussing it. I’m concerned about the confidence of data using markers like this one, mainly when authors states that “weak but distinguishable bands” were scored after electrophoresis. Whats a “weak” band? How to decide when a band is “weak”? Replicates helps to improve the information quality and must be considered during the experimental procedure.

3. As soon as authors described the method for DNA isolation and mention that the DNA is good on quality and in with sufficient amount, I believe that Figure 2 could be omitted. I suggest that authors provide figures concerning DNA isolation as supplemental files.

4. Shannon's information index was presented in Material and Methods section and in Table 3. However, it was not described in the results and neither was discussed. Please, properly present this result and discuss it. Or, remove it from the manuscript.

5. Figure 1 is poor in quality and must be improved. Its resolution is not good enough. We can barely see the province names (e.g., DGD, LS and LSD). We can also barely see the “islands” for the most of sampling localities. I cannot see that Liugong is an island in Figure 1. Are Da Rushan and Laoshan mainland areas? There are too much area in this figure that is not important for the manuscript comprehension. This parts of the figure may be omitted. I don’t think that the whole China map must be present (at least in this size). Emphasizes only what is important for readers understand the main findings.

6. Some parts of the results presentation is only based on showing maximum and minimum values of the estimates. Authors mention measures like averages, but the findings description is focused on the exposition of extreme values. E.g., check lines 174-179. It is important to provide the descriptive statistics of the data and show the general aspects of the genetic diversity. What authors can say about the diversity of populations that didn’t show extreme values?

7. Based on what authors use 0.94 (genetics similarity) to split groups in Figure 4? Despite of the well defined separation of some populations, it is important to mention that the figure scale starts at the 0.88 level. For a more reliable definition of groups it is important to use re-sampling methods in clustering analysis, like bootstrap. Do the authors tried to use this approach? I also suggest using other methods to find out more reliable estimates of number of groups. E. g., Bayesian inference implemented in Structure software (it goes well with dominant data). Or a Discriminant Analysis of Principal Components (implemented in Adegenet package of R). The latter is also preferred to PCoA analysis used by authors due to some particularities of the genetic data. Despite of many references found in literature use PCoA, the percentage of variation explained by the first and second Coordinates is very limited when using molecular markers. It means that with PCoA we loose too much information when we consider only the two dimension plot in analysis. However, if the authors decide to keep using PCoA because they have got high percentage of the variation explained by the two first coordinates, please, show this result.


Specific comments:
1. Line 48: A question: authors evidenced that the species distribution include Jiangsu Province. However this province was not included in the research. Is there an specific reason for not including this province, since it was in the middle of the other sampling localities?
2. Lines 49-51: Please, insert a reference for this statement;
3. Line 52: Please, insert a reference for this statement;
4. Line 68: replace “prime” by “primer”;
5. Line 112: As authors have not used a precise method for DNA quantification, how could they guarantee that they used 30 ng/µL in PCR reactions? If no method like spectrophotometer was used, it is important to mention that this is an approximated measure;
6. Line 128: Does Hs stand for “interspecific” genetic diversity? Isn’t it “intra”? Or better, for authors case, “local genetic diversity”? Please, consider checking it because a single species is being considered in the manuscript.
7. Line 133: R is considered a computational language environment. Is important to mention what kind of analysis was performed using R in methods section. Also, if authors have used specific packages, it is also important to mention them giving their respective references. Include also the R reference (R core team, year).
8. Lines 174-179: estimates of genetic distance and similarity are inversely proportional. Authors don’t have to show both of them. In my opinion it is important to choose one estimate and use it alone so there will be no redundancy.
9. Line 200: Is “PTS” population right? Isn’t it “PTD”?
10. Lines 208-216: Statements presented in these lines look like a discussion. So, authors have to give these information in the proper section.
11. Lines 245-247: Unlike authors affirm, the subpopulation pattern found in Figure 4 disagree with the grouping pattern found in PcoA (Figure 6).
12. Lines 259-264: It is very hardy to compare CDDP marker results with microsatellite (SSR) results when discussing diversity indexes. The letter is codominant and multiallelic, which means that we always expect much more information from SSR than CDDP. A dominant marker like CDDP is interpreted as a bialellic marker and the highest level of heterozygosity expected is 0,5 (or 2pq). For SSR, higher values (e.g., above 0.9) can be found due to the huge amount of alleles found. What I mean is that for a biallelic marker, Hs = 0.5 represents high diversity. However, for SSR, 0.5 of hetherozygosity may represent low level of diversity. Also, comparisons between CDDP and ISSR should be interpreted with caution, since ISSR are potentially not affected by selection. On the other hand, CDDP is developed under the information of functional genes and, naturally, we expect lower levels of variation for these genome regions. Authors suggested in Line 294 that “results of different molecular markers are not fully comparable”, even so they compared results of different markers with no concerning about the main differences among them.
13. Line 309-311: Values of bird propagation distance reported in discussion (480-680 m) are based on unpublished data described in Chung et al., (2002). These distances were poor discussed in Chung et al. The article only cited this values in the INTRODUCTION section and they are not part of a empiric result. Even according to these authors, these values suggests "that seeds are dispersed OVER LONG distances by birds". In addition, this paper brought information about Neolitsea sericea species. I’m not sure if these findings are fully comparable to those found in the present manuscript. I believe that the dispersion by birds should be better explored by authors even with the small gene flow found among localities.
14. Lines 315-317: (a) Angelique (2000) is not listed in References. (b) Whats the specific dispersion information for E. macrophyta? The Angelique (2000) information is for this species? If the ocean is the main responsible for dispersion in E. macrophyta, isn't the species well fitted to this situation? As a reader I need more information to understand this issue because gene flow is a very important issue in your article that must be better explored.